Adobe LLM Optimizer Is Here. Can We Treat AI Search Like a Channel?
Generative search is no longer a future problem. Adobe just shipped the first enterprise application built to solve it.
Adobe LLM Optimizer, generally available since October and now powered by Semrush clickstream data, gives brands a real way to measure, fix, and improve how they appear inside AI-generated answers. Three reasons this matters for companies on Adobe Commerce:
- A measurement layer that ties AI citations and agentic traffic back to engagement and revenue, with side-by-side benchmarking against competitors
- An opportunity engine that surfaces missing schema, FAQs, abstracts, and crawlability issues, then deploys fixes with one click for AEM Sites customers
- Optimize at Edge, which serves AI-friendly HTML to agent traffic at the CDN layer without modifying the origin catalog or CMS
That third capability is the one we’d point to first. Most Adobe Commerce PDPs are JavaScript-heavy SPAs where AI agents see almost nothing on first paint. Optimize at Edge delivers a pre-rendered, structured snapshot to the agent without changing what humans see and without re-platforming. If a buyer asks an AI assistant what the best industrial valves for high-pressure applications are, you want to be the brand that gets named. Right now, most aren’t.
Adobe’s numbers back the urgency. 80% of early-access customers had critical content visibility gaps that blocked AI surfaces from accessing key product information or reviews. Adobe’s own marketing team saw a 5x increase in citations for Adobe Firefly within one week of applying recommendations. The free trial is available to AEM Cloud and Adobe Analytics or Customer Journey Analytics customers, capped at 100 prompts, one domain, and 10 URLs.
For mid-market manufacturers and distributors serious about AI search visibility, this is where the conversation starts.
Talk Strategy
Industry Insight: Shopify Just Moved Into ChatGPT and Claude
On May 4, Shopify launched native connector apps for ChatGPT and Claude. Merchants can now look up orders, update prices, query collection performance, and add products from a photo, all from inside whichever AI assistant they already use.
The number behind the launch matters more than the feature list. 83% of Shopify merchants already use ChatGPT, per Shopify’s internal merchant survey cited at launch. That is roughly 4.6 million stores out of Shopify’s reported 5.6 million active shops. The merchant side of agentic commerce is now eight times more saturated with ChatGPT than the consumer side.
The connector apps follow a deliberate stack of releases:
- Agentic Storefronts (March 2026), which makes Shopify products discoverable by default inside ChatGPT, Microsoft Copilot, Google AI Mode, and Gemini
- The Shopify AI Toolkit (April 2026), an open-source MCP server that gives Claude Code, Cursor, and other coding agents structured access to Shopify’s API, GraphQL schemas, and developer docs
- Native llms.txt files quietly rolling out across stores
For our Shopify Plus clients, this is a workflow shift worth piloting now. Pick one ops person, connect a test store, run a week of price updates and order lookups inside ChatGPT or Claude, and see what falls out of the workflow. AI-attributed orders on Shopify increased 11x between January and November 2025. The operations layer is about to follow.
For our Adobe Commerce clients, this raises a fair question. No comparable native admin connector has launched for Adobe Commerce, BigCommerce, or Salesforce yet. Adobe is investing heavily in MCP and A2A standards through LLM Optimizer, and that work matters. The merchant operating layer, for now, is a Shopify story.
Partner Spotlight: Celigo – The Common Factor in 90% of Successful AI Deployments
AI projects rarely fail at the model layer. They fail because the data is trapped, the systems do not talk, and the workflows cannot reach across silos.
New research from MIT Technology Review Insights, produced in partnership with Celigo and co-branded with Classy Llama, surveyed 500 IT decision makers to find out what separates the organizations operationalizing AI from the ones stuck experimenting. The headline finding: 90% of successful AI deployments share one thing in common. They run on an integration platform.
The report identifies three patterns that show up consistently in the organizations getting AI to scale:
- AI moves out of IT and into every department, with cross-functional teams owning workflows
- Workflows span more data sources and break the silos that used to limit AI to single-use cases
- Teams trust AI with more autonomy today, with growing confidence in greater autonomy ahead
For the manufacturers, distributors, and D2C brands we work with, Celigo is the partner we point to when the conversation turns to scaling AI past the prototype stage. The full report is gated and free to download.